OptimusCloud: Heterogeneous Configuration Optimization for Distributed Databases in the Cloud
OptimusCloud is an online reconfiguration tool designed for cloud database optimization that intelligently adjusts both virtual machine and database management system settings to maximize performance and minimize operational costs.
Researchers at Purdue University have developed a new online reconfiguration tool for cloud database optimization known as OptimusCloud. OptimusCloud jointly tunes VM and DBMS levels to optimize database performance while minimizing the cost. Current cloud technologies using automated decision making, often only working for short and repeat tasks and workloads. Purdue researchers have created an optimal configuration to handle long running workloads. Purdue has achieved 74% better performance, 40% higher throughput, 33% lower cost, and 4.5X lower 99th percentile latency in testing with single server and clusters on three different workloads when compared to currently used systems.
Advantages:
-Cost-effective
-High throughput
-Dynamic
Potential Applications:
-Cloud databases
-Computer technology
TRL: 7
Intellectual Property:
Provisional-Patent, 2020-01-15, United States
Provisional-Gov. Funding, 2020-05-19, United States
Utility-Gov. Funding, 2020-12-31, United States
CON-Patent, 2024-12-20, United States
Keywords: OptimusCloud, cloud database optimization, VM and DBMS tuning, database performance, lower cost, high throughput, low latency, cloud technologies, database optimization, cloud database, Cloud, Cloud Computing, Computational Platform, Computer Programming, Computer Technology, Cost Efficient, Data Processing, Data Science, Data Storage, Database & Information Management, Decision Making, Healthcare, high throughput, Medical IT, optimal compensation algorithms, Optimization, Sequences